arrow_backBack to Blog
Reflection AIDeepSeekOpen-weightModelsInvestmentNvidia

Reflection AI: The $25 Billion ‘DeepSeek of the West’

personLab Desenvolvimento
schedule4 min

Reflection AI, founded in 2024 by Misha Laskin and Ioannis Antonoglou — both Google DeepMind veterans where Antonoglou helped build AlphaGo — is emerging as one of the most ambitious startups in the AI space.

The company is backed by Nvidia, Sequoia, and Lightspeed, having raised $2 billion in October 2025 at an $8 billion valuation. It is now raising again at a $25 billion valuation.

The investment thesis is clear: American governments, banks, defense contractors, and large enterprises want frontier-capable AI with open weights. However, they cannot use closed US labs for sovereignty reasons, and they will not use Chinese open-weight models (like DeepSeek) for security reasons. Reflection is positioning as the third option: American, open-weight, and frontier-scale.

The company has ties to the Department of Energy Genesis Mission and Pentagon AI programs. Its code research agent Asimov has been on a waitlist since early 2026. No public frontier model has shipped yet — the $6.3 billion compute commitment is the most concrete signal that a training run is imminent.

However, Reflection open-weight characterization is contested. Google DeepMind engineer Susan Zhang publicly questioned whether Reflection model release approach qualifies as genuine open source, noting that training data and the full training process remain proprietary.